Learner Simulation on Mistakes
58AccuracyGPT-5-nano
Evaluation results
| Method | Links | |
|---|---|---|
| GPT-5-nano2026.05 | 58 | |
| GRPOBackbone=Qwen3-VL-8B-Instruct2026.05 | 58 | |
| GPT-5.42026.05 | 57 | |
| DITTOBackbone=Qwen3-VL-8B-Instruct2026.05 | 56 | |
| HumanLM-8B2026.05 | 52 | |
| Qwen3-VL-8B-InstructRole=Base2026.05 | 46 |